This checklist is for UK medical oncology trainees, usually ST4 and above, who have worked through most of a question bank and want a defensible way to decide whether they have covered the SCE — not just exhausted a product. The minimum evidence that you have covered the blueprint is not a completion percentage. It is every tumour site and non-site-specific theme attempted under timed conditions, first-attempt accuracy recorded domain by domain, rapidly changing therapies checked against current UK guidance, and a recent unseen mixed block at exam pace. If any of those is missing, you are not finished — you have simply run out of one bank's questions. Read this as a set of things to verify, not a study timetable.
The current SCE Medical Oncology snapshot
The SCE in Medical Oncology is set by the Federation of Royal Colleges of Physicians (MRCP(UK)). It is two papers of 100 best-of-five questions — 200 in total — each paper three hours, on one day, at a Surpass computer-based test centre, with one mark per correct answer and no negative marking. The pass mark is criterion-referenced against the standard of a trainee completing specialty training, so it varies slightly with paper difficulty. Coverage note for the wider market: not every generalist bank covers this specialty — BMJ OnExamination, for instance, publishes SCE banks for several specialties but not for Medical Oncology as of 21 July 2026 — so verify specialty coverage before you buy, and lean on genuine specialists such as StudyPRN, which publishes a dedicated Medical Oncology SCE bank.
The content is drawn from the JRCPTB medical oncology curriculum and samples the whole curriculum by a predetermined blueprint, described rather than published as fixed weightings. The honest position is that every tumour site and every non-site-specific theme is examinable; treat any weighting you have seen quoted as indicative, and note that the common, high-incidence cancers and the cross-cutting themes of acute oncology and systemic therapy carry consistent weight in candidate reports.
Build a blueprint coverage table
Completion counts questions answered; it does not tell you which parts of the blueprint you tested. Build a coverage table with one row per domain and five columns: official weight (indicative), questions attempted, first-attempt accuracy, last reviewed, and confidence. First-attempt accuracy is the number that matters, because re-attempts measure recognition rather than recall — the reason a headline percentage flatters you is set out in our piece on why a Q-bank percentage is not your exam score.
| Domain | Weight (indicative) | Q attempted | First-attempt accuracy | Last reviewed | Confidence |
|---|---|---|---|---|---|
| Breast | High | 65 | 79% | 4 days | High |
| Lung & thoracic | High | 60 | 72% | 6 days | Medium |
| Gastrointestinal (colorectal, upper GI, HPB) | High | 70 | 70% | 8 days | Medium |
| Genitourinary | Medium | 40 | 66% | 2 weeks | Low |
| Gynaecological | Medium | 30 | 63% | 3 weeks | Low |
| Head & neck / CNS / skin | Medium | 28 | 58% | 4 weeks | Low |
| Sarcoma, lymphoma, CUP, NET | Variable | 20 | 52% | 5 weeks | Low |
| Acute oncology & emergencies | High | 45 | 74% | 5 days | Medium |
| Systemic therapy & toxicity | High | 55 | 68% | 1 week | Medium |
| Trials, statistics & governance | Medium | 18 | 60% | 3 weeks | Low |
The figures are an illustrative worked example, not real data. Filled with your own numbers, the table exposes the intersection self-selected practice hides — low volume, old review date and low first-attempt accuracy — which is your true content gap, usually in the rarer tumours and the cross-cutting themes rather than the common cancers you enjoy revising.
Ten domain-level blind spots most likely to stay hidden
Each of these should be verified against exam-specific clinician review before you conclude you are covered.
- Immune-related adverse events — grading and management of checkpoint-inhibitor toxicity across endocrine, gastrointestinal, pulmonary, hepatic, skin and cardiac systems, including when to withhold versus permanently discontinue.
- Acute oncology and oncological emergencies — neutropenic sepsis, metastatic spinal cord compression, superior vena cava obstruction, hypercalcaemia of malignancy, tumour lysis syndrome and raised intracranial pressure.
- Biomarkers and molecular pathology — HER2, EGFR, ALK, ROS1, BRAF, KRAS, mismatch-repair and microsatellite status, PD-L1 and BRCA, and the companion-diagnostic logic that links a biomarker to a therapy.
- Staging and response assessment — TNM nuances by site, performance status, and RECIST response criteria as they appear in stems.
- Systemic therapy pharmacology and toxicity — anthracycline cardiotoxicity, bleomycin pulmonary toxicity, platinum neuro- and ototoxicity, and dose modification in renal or hepatic impairment.
- Rarer tumours and carcinoma of unknown primary — sarcoma, neuroendocrine tumours, germ-cell tumours, mesothelioma and the structured CUP work-up.
- Clinical trials, statistics and research governance — trial design and endpoints (overall survival, progression-free survival, response rate), hazard ratios and confidence intervals, and good clinical practice.
- Supportive and palliative care — symptom control, opioid rotation, anti-emetic selection, malignant bowel obstruction and end-of-life prescribing, taken from the SmPC and UK guidance rather than any single formulary shorthand.
- Cancer genetics and predisposition — Lynch syndrome, BRCA-related cancers, Li-Fraumeni and referral thresholds for genetic testing.
- Survivorship, late effects and prevention — cardiotoxicity surveillance, fertility, second malignancy, bone health, and UK screening and prevention programmes.
Format checklist: the things a bank alone under-trains
A standard bank teaches recall efficiently but under-trains four exam-critical skills. Staging — verify you can assign and use a stage from a described work-up, not just recite TNM tables. Biomarkers — verify you can move from a molecular result to the correct systemic-therapy decision. Rapidly changing therapies — confirm you are revising the current UK position, because SACT and immunotherapy indications shift faster than most bank explanations are updated. Toxicity management — verify you can grade a toxicity and choose the right action (continue, delay, dose-reduce, hold steroids, discontinue), which is one of the most heavily tested practical skills in the paper.
Interpretation checklist
Confirm you have deliberately practised each recurring data type: radiographs and cross-sectional imaging for staging and response; images and described histology; laboratory trends including tumour markers, liver function and the biochemistry of tumour lysis and hypercalcaemia; ECGs where QT-prolonging anti-emetics or tyrosine-kinase inhibitors are involved; and calculations you should do without hesitation — creatinine clearance, carboplatin dosing by the Calvert formula, and body-surface-area dosing. Add the statistics-and-critical-appraisal thread (reading a hazard ratio, a Kaplan-Meier curve and a number needed to treat) and the ethics-and-professionalism thread (consent to systemic therapy, capacity, ceilings of treatment and honest prognostication), both of which recur across the paper.
Recency checklist
Oncology guidance changes quickly, so identify guidance-sensitive topics and record, for each source, its date and jurisdiction. The highest-churn areas at present include immunotherapy indications and toxicity thresholds, biomarker-directed targeted therapy, and antibody-drug conjugates. The jurisdiction column is critical in this specialty: a European or trial-based recommendation may not reflect UK commissioning, so anchor to NICE technology appraisals, the Cancer Drugs Fund position, UK acute-oncology guidance and the SmPC via the electronic medicines compendium, and date every fact. Revising a superseded indication, or a non-UK standard, is a common and avoidable source of confident wrong answers.
Performance checklist: the readiness signals that actually count
Before you stop, confirm five signals. Unseen — recent accuracy is on questions you have not seen before. Timed — you work at roughly 100 questions in 180 minutes with time to review flags. Mixed — you hold accuracy in a blueprint-blind block that jumps between tumour sites and themes. High-confidence errors — you track the questions you got wrong while feeling certain, because these are the dangerous ones a percentage hides. Official-material calibration — you have tested yourself against the Federation's official Medical Oncology sample and online practice test, the closest proxy to the real item style. Retention closes the loop: re-test a domain a week after mastering it, because knowledge that has not survived a spacing interval is not yet exam-ready.
Stop or continue: a decision tree
Use the measured gap, not novelty or unanswered questions remaining in a bank. If a domain is low-volume, low-accuracy and stale, continue new questions there. If accuracy is high but only on re-attempts, consolidate with spaced retrieval so recognition becomes recall. If per-domain accuracy is adequate but your timed, mixed, unseen score lags, simulate full-length blended blocks. If the same misconception recurs despite review — often a toxicity-management or biomarker rule — seek teaching rather than grinding more items. If your unseen scores have plateaued and your errors are careless, rest and taper, because fatigue rather than ignorance is now limiting you. Adding questions in that state lowers your score.
One-page checklist and worked example
Copy this as a one-page gate. You are ready to stop adding new questions when: every domain has been attempted under timed conditions; first-attempt accuracy is recorded per domain and none is both low-volume and low-accuracy; guidance-sensitive topics are dated and jurisdiction-checked against UK sources; you have completed at least one full-length unseen mixed block near exam pace; your high-confidence error log is shrinking; and you have calibrated against the Federation's official sample.
Worked example, illustrative data: a candidate has done 3,500 questions and reports 80% overall, but their coverage table shows 79% first-attempt in breast and 52% in the sarcoma-lymphoma-CUP cluster, last reviewed five weeks ago, and a run of jurisdiction errors where they gave a European rather than a UK-commissioned answer. The 80% is real but inflated by re-attempts and heavy sampling of common cancers. The right action is not to finish the bank; it is to run fresh unseen blocks in the rarer tumours, re-anchor systemic-therapy decisions on NICE, and simulate a mixed paper — the approach formalised in the completion-is-not-coverage matrix.
Where iatroX fits
iatroX is not a specialty-specific oncology SCE bank, and it does not pretend to be. It provides a broad UK, MRCP-level knowledge base with unseen questions and spaced, mixed retrieval, which makes it a clean measurement layer alongside a dedicated SCE bank such as StudyPRN or Licence Medical. Use a specialty bank to build oncology coverage, then use iatroX in standard mode to run an unseen syllabus audit and confirm whether your gains transfer. That is the two-Q-bank rule — a primary bank for coverage, a second unseen bank for measurement — and you can compare options on the iatroX comparison hub.
Three mistakes this checklist is designed to stop
The first mistake is mistaking completion for coverage — finishing a bank while leaving the rarer tumours and the cross-cutting themes of trials, genetics and survivorship barely touched. The coverage table exists to make that gap visible. The second is trusting an overall percentage inflated by re-attempts and by heavy sampling of the common cancers, when first-attempt accuracy by domain is the honest measure. The third is revising oncology as though guidelines were static, importing an international or trial-based answer where UK commissioning differs — which is why the recency and jurisdiction columns matter as much as raw knowledge. Verify against all three and you will know whether you have finished or merely stopped.
Frequently asked questions
How do I know whether I have covered the full SCE Medical Oncology blueprint? You know when your coverage table shows every tumour site and every non-site-specific theme attempted under timed conditions, with first-attempt accuracy recorded and none left both low-volume and low-accuracy. Because the Federation samples the whole curriculum rather than publishing fixed weightings, "covered" means you have deliberately tested the rarer tumours — sarcoma, neuroendocrine, germ cell, carcinoma of unknown primary — and the cross-cutting themes of acute oncology, systemic-therapy toxicity and trials, not only the common cancers. A completion percentage cannot answer this; a domain map can.
Can one question bank be enough for SCE Medical Oncology? One strong specialty bank can carry most of your coverage, but it cannot honestly measure readiness, because once you have seen its questions your accuracy reflects recognition. So one bank is enough for learning and a second, unseen source is needed for measurement — especially in oncology, where you also need to check that your knowledge is anchored to UK practice rather than a bank's international framing. Keep the Federation's official sample in reserve as your calibration standard.
What should I measure instead of my overall Q-bank percentage for SCE Medical Oncology? Measure first-attempt accuracy by domain, your timed and mixed unseen score, your high-confidence error rate, and — specific to this specialty — your jurisdiction-error rate, where you gave a non-UK answer. The overall percentage blends re-attempts and uneven sampling into one flattering figure; these measures show where your gaps are, whether your knowledge holds under exam conditions, and whether it reflects UK commissioning.
When should I stop doing new SCE Medical Oncology questions? Stop when every domain has been attempted under timed conditions, none is both low-volume and low-accuracy, your unseen mixed score sits comfortably above the standard, your jurisdiction errors have largely resolved, and your remaining mistakes are careless rather than knowledge-based. Past that gate, new questions add fatigue rather than information, and your time is better spent on spaced retrieval and full-length simulation.
Which SCE Medical Oncology resource should I use for my weakest component? Match the resource to the deficit. For factual coverage of the rarer tumours, a dedicated specialty bank such as StudyPRN or Licence Medical, plus a focused single-best-answer text, is efficient. For toxicity, biomarker and staging interpretation, drill those item types and calibrate on the Federation's official sample. For UK-jurisdiction errors, re-anchor on NICE and UK acute-oncology guidance. For transfer failures — you know the facts but miss unseen questions — measure and close the gap with a second, unseen bank such as iatroX.
Editorial notes and references
Written by Dr Kolawole Tytler, NHS GP and founder of iatroX. Last checked 21 July 2026. Exam-format facts are from the Federation of Royal Colleges of Physicians and the JRCPTB curriculum; blueprint domains are described rather than quoted as fixed weightings. Coverage note: BMJ OnExamination did not publish an SCE Medical Oncology bank as of 21 July 2026, and the ESMO Examination is a separate European assessment — calibrate against the Federation's official sample. Third-party question counts and prices are vendor-reported and change without notice; verify on the product page. Disclosure: iatroX operates a UK question bank and competes with the specialty banks named here; its role is confined to unseen measurement and cross-specialty knowledge. Corrections via the feedback route on iatrox.com.
References: Federation of Royal Colleges of Physicians — SCE in Medical Oncology and SCE preparation and sample questions; iatroX — why a Q-bank percentage is not your exam score and the blueprint coverage matrix.